改进的蛙跳算法在云计算资源中的研究  

Improved Shuffled Frog Leaping Algorithm in Cloud Computing Resources

作　　者：陈海涛[1] 沈强[2] 

机构地区：[1]中国食品药品检定研究院,北京100050 [2]中国民航信息集团公司,北京100010

出　　处：《计算机与数字工程》2015年第8期1382-1386,1506,共6页Computer & Digital Engineering

摘　　要：云计算条件下的资源调度一直都是目前研究的重点。论文对云计算现状进行了分析,引入蛙跳算法进行资源分配,针对蛙跳算法存在收敛速度快,局部容易陷入最优的情况,首先在蛙跳算法的子群划分中引入人工向量机,其次在算法内部搜索引入自适应交叉概率,改进后的蛙跳算法有效地避免了陷入局部最优,同时缩短了全局搜索和优化的时间。经典函数证明该文算法的性能得到了很大的提高,CloudSim平台表明该文算法可以提高系统处理任务的效率以及云计算资源中合理调度。Resource scheduling under the condition of cloud computing has always been the focus of current research. The current situation of cloud computing is analyzed and introduces shuffled frog leaping algorithm in resource allocation is introduced in this paper. Aiming at that shuffled frog algorithm is easy to fall into local optimum with fast convergence speed, artificial vector machine is introduced into the subgroups of shuffled frog leaping algorithm, then, self-adaptive cross over probability is introduced into the algorithm＇s interior searching. The improved shuffled frog leaping algorithm effectively avoids the algorithm from falling into local optimum and meanwhile shortens the time of global searching and optimurr~ Clas- sic function proves the performance of algorithm in this paper is improved greatly. CloudSim platform demonstrates that al- gorithm in this paper can improve the system＇s efficiency into processing tasks and make resource allocation in cloud compu- ting reasonable and effective.

关 键 词：蛙跳算法 云计算资源 人工向量机 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]